Atlantic's works: Portable layout update

[Atlantic 3279]

The very thing. The official new owner won't know what to do with it anyway.

[Atso]

I'd be only too happy to do so, if somebody could just find me the time to fit in all of my intentions.

In my experience, time is a luxury that is rarely in abundance. On the odd occasions that it is, I am rarely in a position to utilise it effectively!

[Atlantic 3279]

A recent opportunistic addition to the loco stud is a Millholme Model B5, generally neatly built, though not by me, and with some added details. It is certainly tidy enough and runs smoothly enough, as it Is, to be granted running rights until such time as I feel inclined to do a few things to improve it. Power is from a Mashima Can motor via some kind of two-stage gearbox. Early attention will however have to be given to some features of the chassis, because the bogie is on the "good old toy train standard" swinging arm which throws the leading pair of full size (14mm) bogie wheels straight into the cylinders on anything but the gentlest of curves - occasional binding even of a five foot radius is visible. Re-mounting of the bogie either on a reversed swinging arm, or more likely on a pin sliding in a suitably curved transvers slot will alleviate the problem, and a slight compromise on bogie wheel sizes may help further. Those wheels need to be changed anyway as they currently are the 12 spoke variety when the should be ten-spoke. The back-to-backs on the coupled Markits / Romford wheels are a shade narrow too, perhaps as a result of too much cleaning up of wheel boss backs with a file at the time of building, maybe too many cycles of fitting and dismantling - who knows? They are okay running through Peco's "sloppy code 75" point flangeways, but on one or two of my long handbuilt points with true-to-standard narrow gaps around the crossings they bind up slightly. I'll have to whip the wheels off and insert some (paper?) shim washers between the shoulders on the axles and the wheel backs. A couple of side-play limiting washers on the middle wheel set can come out at the same time, as there's pleanty of room for those wheels to have some more sideplay. They have almost none at present and I tend to find that the more flexible the coupled wheelbase is on curves, the better, so long as there's nothing to catch or get knotted up!

Here's the new arrival, middle, compared to a couple of stablemates:

On appropriate stock:

Not the sharpest pictures I've ever taken. I had to borrow the mobile phone from the boss after inserting the batteries wrongly in the stand-by power pack for my own camera, discharging them in seconds after carefully charging them for some time beforehand! Stupid boy.....
I hope the cells have survived the rather hot experience!

[drmditch]

Many thanks for posting. I do like the two-cylinder GC 4-6-0s. (And I have a secret plan for one of their NER contemporaries.)

Are these pictures of your own railway?

[Atlantic 3279]

Those pictures do show the loco on my own loft layout, which doesn't get a lot of use at present owing to time constraints and the usual drawbacks of a loft location. It is quite difficult to select shots that will show enough of a train to be of interest and yet still be capable of simple straight-edge cropping to remove unwanted views of the roof structure, baseboard edges and all the stuff that is necessarily stored in the loft - or is simply there because I haven't had time to tidy up / sell on / throw away!

[Atlantic 3279]

I really should have made a start on test building a new set of P2 valve gear, but over the last two evenings I instead allowed myself to get involved in the business of investigating and solving the minor problem with the running of that B5...

I decided that the first task was to open up the back-to-back gauge of the coupled wheel sets, which I was measuring at 14.2 to 14.4mm on my digital vernier. Certainly my brass gauge block wouldn't go in. I made up some thin washers with square holes to go snugly over the ends of a standard Romford axle and fitted these between the axles and the wheels, opening up the BTB to about 14.7mm, sufficient in principle to easily clear any check rails set at dead-on 14.5mm. In the process I demonstrated something that's probably pretty obvious to anybody who thinks properly, which I obviously did not do at first: trying to dodge the work involved in taking the wheels off from both sides of the loco, by putting all the spacers on one side, is not wise even if this appears to create no big problems with off-centre wheel sets in the chassis. Why? Well, as I found, it is possible to finish up with so little thread engaging when the wheel-nut is re-fitted that the attempt to tighten it partly strips the thread in the nut and the thread at the outer end of the axle The quartering is also less accurately controlled if wheels are not pushed home sufficiently onto the square ends of the axles....

Once I had the spacers correctly arranged, and the previous side-play restricting washers removed from the middle axle, the running of the bare chassis through previous check rail tight spots was much freer. When I put the body back on, the apparently ample clearances previously visible turned out to be something of an illusion. Running of the inverted loco was erratic, with an occasional firework display from the interior of the rear splasher on the side of the loco with insulated wheels and pick-ups, plus occasional lock-up as the con-rod big-ends struck the underside of the running plate. I snarled and set about filing and scraping white metal away from the trouble spots on the underside of the loco, finding that once I had done the initially obvious bits I then had to do some more work within the leading splasher on the pick-up side too. I eventually also coated most of the areas I had scraped/filed back with a thin film of epoxy to give a better guarantee of electrical insulation.

Here's the view under the body before I found the need to do the leading splasher.

I also tinkered slightly with the chassis mounting hole in the swing arm for the bogie, to move the bogie forwards half a millimetre, just to see if that would gain enough clearance to address the previously apparent problem of contact between leading bogie wheels and cylinders.

Further fully-reassembled test running then proved somewhat annoying. Whilst I was fairly sure that the problems with the coupled wheels were eliminated, the loco once again was at times stopping dead on crossings as it took the curved road through points. I took the bogie off again and the problem disappeared. I refitted the bogie and observed closely, concluding that the leading bogie wheels weren't actually touching or even getting worryingly close to the cylinders in some of the dead-stops. It was impossible to see whether the other parts of the bogie, tucked away behind the permanently fitted cylinders and slidebars were making unintended electrical contact, and as my controller doesn't have an ammeter it wasn't immediately obvious whether a short circuit was occurring, or whether, perhaps, it was a mechanical effect such as the bogie or its wheels getting wedged under part of the main frames and lifting slightly the wheels that should have been picking up current. The pick ups, on the coupled wheels only, were all demonstrably functioning well when tested individually. A previous owner of the loco had put lots of little strips of black PVC tape over the frame edges in the cut-outs on the pick-up side of the loco, and on the leading inside edge of the cylinder, which may have helped if electrical shorting was the problem, but might simply have reduced clearances further if mechanical effects were the trouble. The cylinders and slidebars severely restricted access for any contemplated attempt to file out better bogie wheel clearances in the thick brass main frames - why didn't the builder arrange to screw the cylinders/stretcher on to allow for future attention to trouble or simple maintenance jobs, rather than soldering them in place?!

I decided the only practical option, at least as an experiment, without consuming lavish amounts of time, was a switch to smaller bogie wheels. I had some 13mm ten spoke wheels in stock, Markits, with fancy nickel silver tyres too, although the type of spoke isn't really right for a GC loco. Anyway, they are fitted and the stoppages on crossings have ceased. The close-up views below showing the comparison between the bogie wheels, cylnders and smokebox position, both with original large 12 spoke wheels and substitute smaller 10 spoke wheels help to confirm that relative positions are about right but suggest that the frame cut-outs appear huge with the smaller wheels in place. They don't really look that bad until you get close in and down at track level. Maybe 13.5 mm ten spoke wheels would be the ideal compromise, but I'm not sure of availability. That has to be a matter for another time now anyway. I'll paint the sides of the wheel rims black and the loco is useable until I decide to address other features that I would also prefer to improve.

It's interesting that the loco as it was ran so well on Little Bytham. That may be testament to the smooth, level track, the absence of tight flangeways, and possibly something to do with its test route on LB not having included any "sudden" swings to the right or left to negotiate pointwork of any less than around five feet nominal / equivalent radius. Both my own Bardney-like loft layout, and Grantham if the loco ever runs there, require locos to tackle three foot equivalent radius pointwork (or maybe slightly less) in places, without any gentle lead-in. My track is not laid on anything like a billiard table either. It isn't a roller coaster but it does require bogies to have a reasonable amount of vertical freedom of movement. It's all to do with the soft peaty sub-soil on the banks of the river Witham at Bardney you know - nothing to do with my track laying, obviously .

[drmditch]

I haven't built a 4-6-0 - yet! However, many of the problems you describe I do recognise from the 0-4-4,0-6-0s and the 4-6-2 I have built.
In the case of the latter (the A8 shown on my modelling thread), there was an interesting inter-reaction between the rear carrying wheel (which I mounted in a pivoting sub-frame) and the leading bogie. This latter has a curved mounting slot, and one or two other features which you were polite enough to comment on when I posted about them. The loco ran perfectly as a 4-6-0. However, the slight effect of the rear wheels restraining the swing of the fixed wheelbase translated into problems with the bogie,

I wonder whether the tender (depending on how it is attached) could be having a similar effect on your B5 ?

(So annoying of Mr Worsdell, Mr Raven, and Mr Robinson to bequeath us these problems with outside cylindered locomotives!)

PS - I write this with a large degree of hesitation, since you have built so many excellent well-running locomotives!

[Woodcock29]

Hi Graeme
I had similar problems with the bogie when I built my B5 in 1987 with 14mm wheels. I used 12mm Romford wheels in the end as I don't think 13mm were available at that time. On my DJH C2 I used 13mm Gibson. I should try some of those on the B5.

I note that the coping plates seem to be rather a long way back on the tender, they must have been added by the builder given the kit comes with coal rails unless the builder has plated over the coal rails he placed too far back - presumably you'll replace those in due course. The thing I dislike most about the Millholme tenders is the heavy cast band around the top of the sides and back. When I built the B5 I removed that. I've still to add the extra side bit to the water pickup cover on the tender on my B5 though!

Currently I've got out a Millholme Q4 (as provided by the Kit Connection in Spain about 15 years ago) I started many years ago then put aside. I might use a Bachmann tender on that? I'm still trying to work out whether to fit the gearbox to the rear axle or the third which of course would mean it would be visible. I'm probably going to fit a small Portescap I have in which case it has to go on the rear axle with the motor facing to the front as the motor is actually too long if the gearbox goes on the third axle with the motor at the rear. A 4C Porterscap appears to be too wide for the firebox which I've modified to make it the waisted type of firebox.

Regards

Andrew

[Atlantic 3279]

Hi Andrew,

Those coping plates are already on my little list of later jobs, safely tucked into the B5's box. According to whether I eventually decide to keep the loco as 5182 or change it to another class member, the coal plates will either be re-used, moved forwards up to 2mm, or will be replaced by plated coal rails, either externally plated or internally plated. Working out the correct type of tender for the loco and the target year(s) will be a challenge.

I agree about that odd band round the top of the tank sides. It's a bit like an overstated representation of the external framing on a Johnson Midland tender! At least it is less obvious in black than it might be in green livery....

I may get a set of Gibsons, 13mm plain rim ten spoke, for the bogie. I looked in vain for anything 13.5mm diameter on the Gibson list. The offer appears to be 13, 14 or do without. I suppose there's a limit to the number of tyre sizes we can realistically expect Colin Seymour to make.


Dave,

It's certainly always worth thinking about the effect of the tender on the behaviour of the loco, although in this case it does not have a demonstrable critical effect on the problems that were showing up.

[Atlantic 3279]

Having received a parcel almost ten days ago, which to my relief and pleasure contained some etches for the previously missing but necessary parts of the valve gear for P2 model conversions, I finally ran out of excuses today and started some test building of a full set of valve gear including the new parts. The proof of suitability that a test build provides is of course essential if I'm expecting anybody else to accept that the etches are viable.

I must at the outset thank "Atso" of this parish for handling the necessary conversion of my drawings into CAD files of a type acceptable to etchers PPD. He has done rather a good job of it in my opinion, correcting one or two of my misinterpretations of etching "rules" along the way, and introducing a neat idea to assist with one of the laminating tasks, of which more later. Without his valuable help I might still have been waiting or struggling, my previous "partner in etching crime" being rather busy these days and my own assessment of my ability to quickly learn the use of a CAD program being pretty poor!

As we're already well on into the evening here, I won't put up full blow by blow pictures at this stage and fully explanatory construction notes will have to wait for another opportunity, but as evidence that I've actually done something, I offer the following:

Part of the new etch, which as a full sheet includes several sets of parts for the P2 gear and some other bits handy to me filling up otherwise wasted space and wasted metal:

Some parts with holes drilled out to final size, separated from the etch, and cusps filed off. The fold-up item forms the foot or drop link required to allow re-use of the original Hornby "Cock o' the North" plain crosshead (Lentz rotary cam valve gear) on a model converted to carry Walschaerts valve gear. The forked part of the link neatly traps a raised feature on the face of the crosshead when used correctly, forcing the link to stay vertical as it should, rather than flopping back and forth spoiling the action of the valve gear. I adopted this idea in order to eliminate the need to either build unrealistic valve gear or to fit new crossheads, such as a pair of (multi-part) Branchlines items that I had originally planned to use, but which might not necessarily fit the re-used Hornby slidebars.

All I can say at this stage is that these parts for the front of the valve gear go together exactly as I had envisaged and fit the existing "slimmed" resin cylinders for my Lord President model. The drop link or crosshead foot seems to be held very firmly indeed.

The next item to test is the adaptor plate for the motion bracket stretcher (the latter originally designed for A2 models) and then the crucial job of proving that the rear parts of the valve gear and the new front elements go together without snags. More to follow as the week goes on....

It's good to be building from etches again, rather than banging my head against the wall trying to find time for umpteen overdue tasks .

[mick b]

Nice work !! look forward to fighting with Mons Meg in due course !!.

[Atlantic 3279]

I'm not progressing at lightning speed as I am writing full instructions as I go, while it's all clear in my mind.

For those who might feel that valve rod/spindle is indispensable, although such things cannot generally be seen on the finished model and the radius rods are eventually secured in such a way that the combination levers are supported:

Motion bracket stretcher adaptation piece prepared by soldering two lengths of 1/16" sq brass bar along it edges:

This now flipped over and tried over the relevant part of the chassis block to ensure that it will drop on, with a little bit of wiggle room available:

Next soldered to the top face of the A2/2 and A2/3 motion bracket stretcher, thus:

Two part-depth piercing saw cuts run into the original stretcher on the underside to show the intended (later) cutting-out of the central part:

Avoiding any risk of heat spreading from other significant joints and making the whole thing fall apart again, with loss of alignment as a result, I decided to fold up and solder the outer bracket parts then added the inner bracket / inner slidebar pieces before completing the cuts. That phase of soldering completed, and the lower stretcher sawn out centrally, gave me this:

With the top block lifted sufficiently from the main chassis and the inner slidebars slotted carefully into place behind the crossheads and other gubbins, I was able to lay the completed "core assembly" in roughly its final position:

And this emphasizes the reason for the strange shape of the adaptor piece - it allows the motion bracket to be fitted to, or removed from the chassis as many times as may be necessary without any need to disconnect and re-connect the wiring from the pick ups:

I hope people are not put off by that one instance of the need to saw metal carefully. It isn't dreadfully difficult - it's easy in fact by comparison with recent attempts to discuss the Locomotion / Rapido Stirling Single model over on RMWeb. My method doesn't require a substantial supply of body bags to facilitate the clear up process......

[mick b]

The most pathetic part of this its only a EP photo that has been posted ,probably not a lot like the final version. Body bags ......... new supplies arriving anytime now, it not on the Sterling thread , the large one is laying down the rules , I am sure they can find another thread to wibble one soon !! .

Strange how the same people, come up time and time again on most of the RM Web handbag threads and 90% never post a picture of any kind of model ?? perhaps too busy expelling hot air



Valve gear looks excellent , nice design .

[Atlantic 3279]

I'm glad to report that the remainder of the valve gear is now built up, basically as per the original A2 instructions, and is working nicely - done with the fancy two layer expansion links too, to prove full viability. I just have the etched reversing rod parts to test, some minor tweaks to make to the instructions, and some costs to estimate, then we might at long last be "good to go".
More pictures soon.

[Atlantic 3279]

As promised. Firstly, a full set of working valve gear in place on a P2 chassis with the top block screwed back down to clamp the motion bracket stretcher in place. I haven't bother to file a shallow slot anywhere to accommodate the thickness of the stretcher, which, as you can see, now prevents the front of the top block from nipping down fully onto the main chassis block and the rear of the cylinder stretcher. It doesn't seem to matter. There is room inside the bodies for the slightly raised block, and were the cylinder stretcher to be a less than secure fit in its slot it would be simple enough to put a sliver of plastic or a metal shim under the nose of the top block to impart clamping pressure to that stretcher.

The reversing rod:

Our forum's resident N Gauge CAD expert has designed in a handy little aid to accurate lamination of the two layers of the reversing rod.

That long half etched line between the two layers of the rod allows you, with appropriate care, to fold that part of the etch double, with the fold line and the half-etched pips on the reversing rod itself on the outside face / faces, accurately lining up and holding the two layers of the rod ready for soldering. Two etched holes are also provided for additional (or alternative) means of alignment. I started by tinning (thinly) the faces of the rod that were to be joined together, before trying to fold anything. Attempting to fold the etch along that long line can be "interesting" because the metal would obviously prefer to fold in etched gaps around the parts, especially near the narrow nose of the reversing rod. Either very effective bending bars with a good clamping action, or very cautious forming of the fold bit by bit with pliers (as I did, but strenuously avoiding the creation of distortions), or careful deepening of the half etch line prior to folding may be the ways to achieve success here. The etch frame could even be cut along the half etch line (again avoiding distortion) and just the two holes used for lining up the layers. Anyway, with the two layers accurately and closely superimposed, pressed down on a flat surface with a little piece of wood for instance, a little more flux and a slow wipe over with the soldering iron joins the parts nicely. Heat loss into the etch frame through the few slim tags is minimal. The tags can then be cut and the piece cleaned up.

The strips designed to form the supporting brackets can be folded up into the shape shown in the example in foreground of this image, then soldered onto the laminated rod in the positions marked by the etched pips on the rod.

Three half etched notches are provided on each strip to guide and aid the folding process, although I have discovered that my calculations were slightly adrift in respect of one of the positions. The two closely spaced notches mid-way along the strip, that allow it to form around the base of the reversing rod are correct, the folding here closing up the notches, as is conventional. The fold at the other single notch around a quarter of the way along each strip was always intended to be made the other way, I.e. opening up the notch, but test building has revealed that this notch isn't in quite the ideal place. In order for the folded bracket to support the reversing rod, as intended, in the proper position below the running plate, a new and slightly different fold position is required, and notching the strip with a triangular file may assist.

This is mildly annoying, but not in my view a serious enough matter to justify the cost of revised etches.

The idea of the "odd legs" on those brackets is that the straight, vertical one (trimmed to length if necessary so that it doesn't project through) will go up into a hole drilled in the running plate, the one bent at right angles will be glued (epoxy?) to the underside of the running plate. The two legs could of course both be left straight and pushed up into holes in the running plate. The "odd legs" arrangement does offer the possibility of holding the reversing rod against the edge of the running plate as shown below, so that the bracket positions can be marked off on the underside accurately ready for the drilling of suitable holes....